Yu-0312

驗算報告、論文、財報或簡報裡的圖表與統計值——用原始數據重新計算文件聲稱的每一個數字（平均、斜率、R²、CAGR、百分比變化、誤差、活化能…）， 抓出原文算錯的地方，如實標註待核而不擅自改數字、不捏造誤差、不指控原作者。 Use this skill whenever the user asks to verify, recompute, sanity-check, audit, or "double-check the math" on a chart, figure, table, or claimed…

把凌亂的文件（會議逐字稿、多場會議混檔、實驗／數據報告、論文、財報）變成可直接上台的設計感簡報的工作流。 Use this skill whenever the user wants to turn a document, meeting notes, transcript, report, spreadsheet, or raw data into slides / a presentation / a deck — 包括他們說「做成簡報」「整理成投影片」「幫我做一份簡報」「把這份報告變簡報」 「turn…

Universal visual direction engine for AI image, design, and layout work: model-agnostic decision pipeline (intent, analysis, visual language, planning, recovery/refinement) plus swappable adapters for GPT Image, Flux, Ideogram, and generic image backends. Use…

Recompose objects from an uploaded photo into polished editorial, kawaii-clean, or kawaii-story images while preserving object identity. Use when Codex should rearrange food, drinks, cameras, books, tableware, furniture, decor, or products into deliberate…